A Comparative Survey on Multi-Language Sentiment Analysis Methods

Firm leaders can utilize sentiment analysis to track public opinion on products, services, brands, and news stories. These dashboards track good, negative, and neutral Twitter comments from millions of users. Because these technologies can't automatically recognize sentiment fluctuations, decision-makers struggle to reach the right conclusions. The authors assessed a pool of potential Sentiment Analysis classifiers for short texts and identified the best performer. The authors provide Filtered-LDA, a more accurate approach for measuring Twitter users' moods. The method uses a hierarchical network of LDA Models with adjustable hyperparameters to discover public opinion shifts. Using a Topic Model with a high Coherence Score, Emerging Themes and weary themes can be distinguished. Finally, a Twitter dashboard for reasoning about sentiment organizes and provides the most prevalent tweets for each probable explanation.